« ETech or bust   |   Main   |   Networked objects at ITP as art… Etech2005 »

Wed morning at ETech2005

So we’re down here at ETEch - I paid for the day pass, since a few people asked. I was planning on SXSWi, and an Etech day pass was essentially the same price. It’s an expensive conference and since I’m untethered to corp entities, my evalutation is that of a private citizen. It’s interesting how a pragmatic and flexible approach can change the context: a quick flight down this am got us here in good order and we fly back tonight.

Cory’s talk was good but his is readable online (and was indeed read aloud). I like Cory better in print anyway - writing for speaking always has a different rhythm.

The swarming presentation is online now from SwarmStream - it’s funny to hear how this actually connects to work the US folks at NEC (with whom I sat rather than worked directly) are doing with their fault tolerant servers and approach to redundancy. I’m glad to see that some of my work on more boring enterprise software stacks, with areas like Content Delivery Networks, is still relevant to stuff going on here (of course it is, but it it’s different than what you see looking out the IDC libray window).

Now we have Wikipedia and the Future of Social Computing with Jimmy Wales, President, Wikimedia Foundation, Inc. They’ve got 499, 388 English language articles, German is third at 209K, and Japanese is third at 106,682 articles and just tied the NYT in page views (~500M pageviews monthly). Lots of other stuff.

Now folksonomies are on, with hey - the del.ico.us guy, Josh Schacter (who we were just chatting with out in the hall, but I didn’t connect him!). He’s apparently just a stockbroker type who scratched his own itch and created a great tool. He’s supposed to be very nonchalant about the fuss, and happy in his own vocation (and therefore del.ico.us tends to languish and go up and down). That’s just what people who know have told me. But here the man himself speaks: He said del.ico.us started just from a text file, just collecting to post on a website, 20k odd links. Somewhere there he started putting hashmarks and text to grep them out later, his wireless links etc. The web version came later, but still single user, for him. It was less cateogorization, but just grouping (which for me is the key issue - let the structure emerge). Then he made it massively multiplayer. What he found interesting was about the behavior around the tags with the users, as they triangulaed around both the quality of the url/doc and the context of the user combined. He said that people use tags for all sorts of things. Tags for groups working together, tags for workflow and rss to feed stuff around. Ah, most interestingly, Josh notes that wiki’s people fight for space, there is a limited namespace and taxononmy is much more rigorous (my edits), and that del.ico.us is much more individual - there’s no agreement necessary. That’s a good distinction to make.

Marc Canter asked if we could connect tags together, the tags we’ve left lying around in the various services. I’ve been looking at this as part of my side projects, drawing on the whiteboard and talking to folks like Michael Eakes of Rojo. Something like an exportable opml that I can own as my microcontent and bring into these different system. Stewart B says that Technorati has most of this now (they have it, but it’s my microcontent, and in Technorati it’s not well linked to me). Josh says we’re already starting to see this, different axes of what tagging, why tagging, and what happens. Flickr is mostly my stuff for my purposes. Technorati is tagging my stuff for other people’s purposes. Del.ico.us is tagging other people’s stuff for my use. Clay said that the api’s are there and that the answer (totally agree) is that it should be a local activity with an upload to servers. He also brought in the idea of the remix culture, which is a good connection. But for me this is all about my microcontent working for me along all the axes of use and the variable functionalities and supported use cases of the various tools.

Josh was talking and I was goofing off and now he’s talking about a sort of dashboard interface of my tags, commonly associated (popular?) tags (perhaps associated with the items (what other people would tag them) and then a third component. I’ll have to find Josh again, because I violently agree with all that he’s saying.

Anthony Eden (who wrote TagSurf just asked how we would tie in together all the tags. Technorati is searching thru 3 services, Flickr, Furl and blogging. Josh said are you tagging for your own purposes or for others? There’s nothing to keep people from misbehaving if not - the whole meta tag padding issue. (Defefinitely look online for other notes on this session - it is VERY interesting). I’ll post this, and add update links and possibly that third thing from Josh if I remember to ask him.